33 results about "Auto segmentation" patented technology

Disclosed is a computer-implemented method which encompasses registering a tracked imaging device such as a microscope having a known viewing direction and an atlas to a patient space so that a transformation can be established between the atlas space and the reference system for defining positions in images of an anatomical structure of the patient. Labels are associated with certain constituents of the images and are input into a learning algorithm such as a machine learning algorithm, for example a convolutional neural network, together with the medical images and an anatomical vector and for example also the atlas to train the learning algorithm for automatic segmentation of patient images generated with the tracked imaging device. The trained learning algorithm then allows for efficient segmentation and / or labelling of patient images without having to register the patient images to the atlas each time, thereby saving on computational effort.

Systems and methods are provided for multi-schema analysis of patient specific anatomical features from medical images. The system may receive medical images of a patient and metadata associated with the medical images indicative of a selected pathology, and automatically classify the medical images using a segmentation algorithm. The system may use an anatomical landmark detection algorithm leveraging Deep Reinforcement Learning (DRL) techniques to automatically locate one or more anatomical landmarks associated with the patient specific anatomical feature within the medical images. A 3D surface mesh model may be generated representing the patient specific anatomical features including the located one or more anatomical landmarks. The located one or more anatomical landmarks may be used to guide placement of a 3D model of a medical device that may be fused with the 3D surface mesh model to generate a patient specific 3D model of the medical device.

A deep-learning based framework to assess the quality of medical image auto-segmentation is described. According to an example, a computer-implemented method comprises receiving segmentation masks generated, via one or more segmentation models, from medical image data depicting an anatomical region of a subject, wherein each of the segmentation masks depicts a different anatomical structure of a set of different anatomical structures included in the anatomical region. The method further comprises generating reconstructed versions of the segmentation masks based on application of a multi-channel reconstruction model to the segmentation masks, wherein the reconstructed versions correspond to optimized versions of the segmentation masks. The method further comprises determining an assessment of quality of the segmentation masks based on comparison of the segmentation masks to the reconstructed versions, generating output data regarding the assessment of quality, and rendering the output data via an electronic output device.

Various embodiments include methods for generating patient-specific heart and thorax models using anatomical landmarks and segmentation data, optimized through the application of trained neural network models. Three-dimensional medical imaging data may be processed by a trained neural network to automatically segment and isolate the heart, blood cavities, and thorax to extract feature maps from the segmented images. Anatomical landmarks, such as the heart apex and valve centers, are identified and the alignment of heart axes is verified. Reference heart and thorax models are selected and adapted to fit the patient-specific landmarks through scaling, translating, and rotating. Best adapted heart and thorax models may then be used for conducting one or more medical procedures. Neural network models, trained on historical patient data sets, may be refined through machine learning from new patient data, thereby improving accuracy.

The present disclosure provides systems and methods for multi-modal analysis of patient-specific anatomical features from medical images. The system can receive a medical image of a patient and metadata associated with the medical image indicative of a selected pathology, and automatically classify the medical image using a segmentation algorithm. The system can use an anatomical feature identification algorithm to identify one or more patient-specific anatomical features within the medical image by exploring an anatomical knowledge dataset. A 3D surface mesh model representing the one or more classified patient-specific anatomical features can be generated, such that information can be extracted from the 3D surface mesh model based on the selected pathology. Physiological information associated with the selected pathology for the 3D surface mesh model can be generated based on the extracted information.

The invention discloses an SEEG connection analysis method and system under brain region division, and relates to the technical field of computer vision, and the method comprises the steps: fusing MRI and CT images of a patient, constructing a three-dimensional brain region model based on a Destrieux atlas and an automatic segmentation algorithm, carrying out multi-modal registration to calibrate the spatial position of an SEEG electrode, and generating a primary electrode file with an anatomical tag; a virtual electrode point is generated by adopting bipolar combined guidance, and the affiliation of an electrode brain region is dynamically corrected in combination with spatial registration and prior implantation information, so that a three-level tag file is formed. According to the method, SEEG electrode pairs are grouped, sorted and identified according to brain regions, brain region-level summarized signals are output, connectivity analysis of cross-brain region electrode pair groups is supported, and spatial interpretability and analysis precision of SEEG data in brain network research are improved. According to the method, accurate grouping of SEEG data according to anatomical brain regions is realized through multi-modal image fusion, virtual electrode construction and a prior information guided brain region affiliation correction technology.

CT angiography can be used to obtain a 3D anatomical model of one or more blood vessels in a patient, while simultaneously obtaining 2D images of the vessels based on perspective. The 3D model can be registered with the 2D images based on contrast agent injection sites identified on the 3D model and / or in the 2D images. A fused image can then be created to depict the overlaid 3D model and 2D image, for example, on a monitor or via a virtual reality headset. Injection sites can be determined automatically or based on user input, which may include bounding boxes drawn around the injection sites on the 3D model, selection of automatically segmented regions in the 3D model, etc.

The present invention relates to an artificial intelligence-based dentition matching and jaw bone segmentation method, system and medium, including the following steps: S1, constructing a dentition matching algorithm under constrained movement to obtain an objective function for matching a digital dentition model with the dentition of a skull model; S2, minimizing the objective function to obtain an automatic matching method for the upper and lower dentitions with the dentition of the skull model, and the translation and rotation methods of the upper and lower dentitions. The jaw bone segmentation method includes: using this dentition matching method to replace the tooth part on the three-dimensional skull model with the upper and lower dentitions of the digital dentition model; then automatically segmenting the condyles of the three-dimensional skull model to complete the automatic segmentation of the jaw bone of the three-dimensional skull model. This application constructs an intelligent dentition matching method and jaw bone segmentation method for orthognathic surgery, which can automatically complete the important links of jaw bone segmentation and dentition matching in digital orthognathic surgery design by computer, so as to reduce the workload of clinicians and improve work efficiency.

This invention discloses a medical image segmentation method and system based on local Moran's index residual blocks, belonging to the field of intelligent medical image segmentation technology. By introducing multi-scale local Moran's index features, this invention effectively quantifies the spatial clustering patterns of key anatomical structures in shoulder MRI images, achieving accurate characterization of eight types of structures. A residual Moran's feature module is deeply embedded in the UNet architecture, and spatial feature utilization is enhanced at each level of the encoder-decoder, improving the model's adaptability and generalization ability to complex shoulder anatomy, making the segmentation results more robust and reliable in clinical applications. The constructed segmentation model can achieve end-to-end automatic segmentation of shoulder MRI images without additional spatial statistical preprocessing, offering advantages such as convenient deployment and high computational efficiency. It is suitable for rapid clinical diagnosis and surgical planning, and has significant application value in improving the intelligent diagnosis and treatment of shoulder joint diseases.

Disclosed is a computer-implemented method which encompasses registering a tracked imaging device such as a microscope having a known viewing direction and an atlas to a patient space so that a transformation can be established between the atlas space and the reference system for defining positions in images of an anatomical structure of the patient. Labels are associated with certain constituents of the images and are input into a learning algorithm such as a machine learning algorithm, for example a convolutional neural network, together with the medical images and an anatomical vector and for example also the atlas to train the learning algorithm for automatic segmentation of patient images generated with the tracked imaging device. The trained learning algorithm then allows for efficient segmentation and / or labelling of patient images without having to register the patient images to the atlas each time, thereby saving on computational effort.

A system and method are provided for multi-scheme analysis of patient-specific anatomical features from medical images. The system may receive a medical image of a patient and metadata associated with the medical image indicative of a selected lesion, and may automatically classify the medical image using a segmentation algorithm. The system may identify one or more patient-specific anatomical features in the medical image using an anatomical feature identification algorithm by searching an anatomical knowledge dataset. A 3D surface mesh model may be generated representing the one or more classified patient-specific anatomical features such that information may be extracted from the 3D surface mesh model based on the selected lesion. Physiological information associated with the selected lesion for the 3D surface mesh model may be generated based on the extracted information.

The invention relates to a liver surface landmark segmentation method based on a double-flow grid convolutional neural network, and the method comprises the steps: constructing a data set which comprises liver grid data and corresponding marks; building a double-flow grid convolutional neural network TSMCN which comprises two parallel flows, namely an edge processing flow E-stream and a point processing flow P-stream, and is used for extracting geometric representations from grid edges and coordinate points respectively; training and optimizing the TSMCN by using the data set to obtain a liver image segmentation model; and inputting the current liver grid data into the liver image segmentation model, and outputting to obtain a segmentation result corresponding to the key anatomical region. Compared with the prior art, the method breaks through the limitation of a traditional single-flow network, can learn geometric features from two key perspectives at the same time, provides multi-dimensional information support for accurate segmentation, achieves accurate and automatic segmentation of key anatomical areas on the liver surface, and is beneficial to improving the effect of augmented reality (AR) guided liver operations.

This invention discloses an automatic identification method for pelvic organ prolapse based on deep learning and the PCL line, comprising: S1: acquiring and preprocessing medical images of the pelvic floor region, and simultaneously annotating the bladder, uterus, rectum, and bony anatomical structures to construct a training dataset; S2: constructing and training a multi-objective deep learning segmentation model based on the training data to achieve automatic segmentation of the bladder, uterus, rectum, pubic symphysis, and the two vertebrae from the coccyx; S3: extracting the lowest edge point of the segmented pubic symphysis region and the highest edge point of the two vertebrae from the coccyx region to construct a set of bony key points; S4: connecting the bony key points to automatically generate the pelvic floor reference line (PCL line); S5: analyzing the spatial positional relationship of the bladder, uterus, and rectum relative to the PCL line to obtain the downward displacement feature information of the pelvic organs; S6: automatically determining the degree of pelvic organ prolapse based on the spatial positional relationship. This invention achieves automated identification of the degree of pelvic organ prolapse, reduces subjective errors caused by manual measurement, and improves the consistency, accuracy, and efficiency of the assessment.

Provided are a method and an apparatus for auto segmentation using a bounding box. A method for auto segmentation using a bounding box according to one embodiment of the present disclosure comprises receiving a first object image including an object labeled with a bounding box, which is a pre-learning target, learning a segmentation model by classifying an object and a background from the bounding box of the received first object image, and segmenting an object from a second object image, which is an identification target, using the learned segmentation model.

This invention discloses an automatic bone structure segmentation method, system, electronic device, and storage medium. The method utilizes a pre-trained multi-level, multi-task model group for subdividing the bone structure of a single vertebra, achieving automatic segmentation and extraction of bone structures. This reduces manual intervention and solves the problems of low efficiency and unreliability associated with manual annotation and segmentation of bone structures in existing technologies, laying a solid data foundation for subsequent clinical surgical planning. Furthermore, based on clinical indications and planning requirements, the method uses pre-trained primary and secondary cutting models to further subdivide the bone structure of a single vertebra. The multi-level, multi-task fine segmentation model group strategy, segmenting substructures from coarse to fine, can significantly improve segmentation accuracy.

The invention relates to a liver grid anatomical landmark segmentation method based on nested resolution grid-graph convolution, and the method comprises the following steps: S1, obtaining high-resolution liver grid data, and obtaining low-resolution grid data through random edge sampling; s2, taking the low-resolution grid data as an input feature of a dynamic graph convolutional network to obtain a low-resolution anatomical landmark preliminary segmentation result; s3, expansion mapping and erosion mapping processing are carried out on the preliminary segmentation result; and S4, based on the expansion mapping data and the erosion mapping data after mapping processing, in an anatomical landmark refining network based on MeshConv, outputting a high-resolution segmentation result through attention fusion. By adopting a nesting resolution grid-graph convolution architecture, the method aims to solve key problems existing in automatic segmentation of liver 3D surface anatomical landmarks in an existing method, so that the accuracy and robustness of segmentation are improved.

The invention provides a scoliosis minimally invasive surgery auxiliary system for recognizing vertebral body morphological changes based on AI, and relates to the technical field of medical instruments and artificial intelligence. The system comprises a multi-mode image input module for acquiring and preprocessing spine image data; the vertebral body segmentation and three-dimensional reconstruction module is used for automatically segmenting a vertebral body and reconstructing a three-dimensional surface model; the AI vertebral body micro-morphological change identification module is used for quantitatively analyzing the micro-morphological change of each vertebral body; the scoliosis comprehensive evaluation module is used for automatically measuring a Cobb angle, centrum rotation and trunk offset in combination with the tiny morphological change and the overall spine force line; the minimally invasive surgery planning and navigation module is used for automatically designing a minimally invasive surgery scheme according to the individualized centrum micro morphological characteristics, planning a filling part and material consumption and distribution, and providing intra-operative navigation support; the minimally invasive surgery scheme guides a doctor to fill, rest and correct the asymmetric part of the vertebral body by using a filling material under a minimally invasive condition. And on the basis of AI identification, minimally invasive open surgery is realized.

The application discloses a SEEG connection analysis method and system under brain region division, and relates to the technical field of computer vision, and the method comprises the following steps: constructing a three-dimensional brain region model based on a Destrieux atlas and an automatic segmentation algorithm by fusing MRI and CT images of a patient, performing multi-modal registration to calibrate the spatial position of SEEG electrodes, and generating a first-level electrode file with anatomical labels. A virtual electrode point is generated by adopting a bipolar connection, and the brain region attribution of the electrode is dynamically corrected by combining spatial registration and prior implantation information to form a three-level label file. Accordingly, SEEG electrode pairs are grouped, sorted and identified according to brain regions, brain region-level summary signals are output, and the connectivity analysis of electrode pairs across brain regions is supported, so that the spatial interpretability and analysis accuracy of SEEG data in brain network research are improved. Through the multi-modal image fusion, virtual electrode construction and prior information guided brain region attribution correction technology, the application realizes the accurate grouping of SEEG data according to anatomical brain regions.

An anatomy of interest, such as the aortic arch, is automatically segmented from magnetic resonance imaging (MRI) data. The segmented anatomy volume is then processed to track or otherwise determine a centerline of the anatomy volume. Based on this centerline, quantitative measures of the anatomy of interest can be computed. The resulting segmented anatomy volume, centerline, and / or quantitative measures can then be displayed to a user or stored for later use and / or processing. The segmented anatomy volume, centerline tracking and quantitative measures can be monitored to determine longitudinal deformations in the subject relative to previous measures from the subject, or relative to normative measures from an age-matched, sex-matched, race-matched, and / or disease-matched cohort.

The invention provides a three-dimensional total hip replacement prosthesis type selection method and device based on reinforcement learning, a medium, a program product and a terminal, and the method comprises the steps: obtaining the medullary cavity form and prosthesis three-dimensional data, screening and training a prosthesis model, carrying out the structural division, limiting calculation and geometric simplification, and generating polyhedral data, and finally, generating a model selection model on a Unity platform by utilizing reinforcement learning, and selecting an optimal prosthesis model for replacement. According to the method, the regions can be automatically segmented in the medullary cavity, the anatomical mark points can be identified in an auxiliary manner, intelligent detection and planning of the THA operation are realized, the labor cost is greatly reduced, the diagnosis efficiency is improved, and personalized clinical requirements are met. The optimization algorithm improves the accuracy and speed, reduces the consumption of computing resources, and can provide efficient surgical planning even in areas with limited medical resources.

The invention discloses an automatic segmentation method for a solid waste landfill functional area, and relates to image recognition, the method comprises the following steps: collecting a remote sensing image, preprocessing the remote sensing image, and constructing a data set; and performing time sequence segmentation processing on the remote sensing image in the data set through an improved RSIE-SegFormer model so as to obtain a segmentation result used for representing the solid waste landfill functional area. According to the method, remote sensing images of landfills in multiple regions are collected and preprocessed based on a remote sensing platform, the remote sensing images are fused with a disclosed reference data set, a landfill functional region data set containing multiple categories and multiple time phases is constructed, and the problem of scarcity of the data set in the field is effectively relieved; the improved RSIE-SegFormer model is introduced into time sequence segmentation of the landfill functional region, so that the feature extraction capability of the landfill functional region is enhanced, and the segmentation precision of boundary details is remarkably improved.

The invention provides a rapid and automatic tooth three-dimensional model segmentation method and equipment based on deep learning, and the method comprises the steps: carrying out a data collection process: collecting and preprocessing three-dimensional tooth point cloud data, and carrying out the reconstruction of a triangular mesh and the separation of an upper jaw and a lower jaw, preprocessing to obtain a grid model of an upper jaw and a grid model of a lower jaw; the rapid segmentation process comprises the step of segmenting the preprocessed model based on an edge refining cascade network, and the edge refining cascade network comprises a coarse segmentation branch and an inter-tooth refining branch which are connected in series; and in the post-processing process, a result obtained in the rapid segmentation process is optimized. According to the invention, accurate and rapid segmentation of the tooth grid model can be realized.

The application discloses a kind of based on real virtual constraint environment and biomechanics analysis's impacted wisdom tooth minimally invasive extraction intelligent surgical planning method and system, belong to oral medical computer-aided surgery technical field, including: obtaining the oral cavity data of patient oral cavity maxillofacial part, and utilize deep learning model to carry out automatic segmentation to oral cavity data, generate oral cavity maxillofacial three-dimensional model;Anatomical constraint parameters are extracted based on oral cavity maxillofacial three-dimensional model;Based on anatomical constraint parameters and operating space constraint parameters, the optimal global tooth extraction strategy is determined by strategy decision model;According to the global tooth extraction strategy, path planning is carried out, and the optimal extraction path for the whole or each tooth body segment after tooth is divided is generated;The optimal extraction path is converted into robot executable operation instruction sequence.Through three-dimensional modeling, anatomical constraint extraction, strategy decision and path planning, the intelligent surgical planning of impacted wisdom tooth minimally invasive extraction is realized.

The application is suitable for the technical field of action recognition and prediction, and provides a human action recognition and prediction method based on motion time sequence feature coding, which comprises the following steps: establishing a human skeleton model, constructing an action sequence based on the human skeleton model, segmenting the action sequence according to an automatic segmentation model to obtain a sub-action sequence, extracting key frames from the sub-action sequence according to a clustering algorithm, encoding and assigning the key frames to obtain an encoding table of the key frame corresponding digital sequence, re-distributing the weight proportion of ordinary frames and key frames based on an attention mechanism, constructing a human action recognition model according to a long short-term memory recurrent neural network and the attention mechanism, and obtaining a test action sequence and an action recognition result according to the human action recognition model. The application can effectively improve the recognition and prediction efficiency of the model by using key frames to describe the motion sequence for the behavior action sequence of the human body.

A method for segmentation of a cardiac myocardium in one or more images of a subject includes receiving at least one image of a heart of the subject, a segmentation of at least one heart structure, and an identification of a right ventricle insertion point, providing the at least one image of a heart of the subject, the segmentation of the at least one heart structure, and the identification of a right ventricle insertion point to a segmentation model, and generating, using the segmentation model, a subject specific seventeen segment myocardial contour model.

The application provides a kind of artificial intelligence-based tricuspid annulus accurate positioning and plane inclination automatic calculation method, comprising: obtaining and preprocessing three-dimensional medical image;Based on deep learning model, key structures such as right atrium, tricuspid valve and right ventricle are automatically segmented;Extraction of leaflet junction and annulus reference point, construct initial annulus curve;Through path optimization and resampling, improve curve continuity and point distribution uniformity;Using iterative strategy to optimize the spatial distribution of annulus point set, so as to fit the surface characteristics of valve leaflet;Determine the annulus plane and calculate the normal vector by using mathematical fitting method, and further obtain the included angle with horizontal plane.The method realizes the automatic identification and inclination quantization of tricuspid annulus, improves the accuracy and clinical applicability of morphological measurement, and reduces manual error.

Provided herein are systems and methods for automatically segmenting a 3D model of a patient's teeth. A patient's dentition may be scanned. The scan data may be converted into a 3D model, including a graph-based representation of the 3D model. The graph-based representation can be input into a machine learning model to train the machine learning model to segment the 3D model into individual dental components. Trained machine learning models can also be used to segment graph-based representations of a 3D model of a patient's teeth.

The application relates to the field of artificial intelligence medical image processing technology and discloses a three-dimensional intelligent jaw bone absorption rate dynamic change monitoring method based on CBCT. CBCT data is used for three-dimensional image analysis, a deep learning model is used to automatically segment teeth, tooth enamel, maxillary bones and mandibular bones, the enamel-dentin junction and the top base of periodontal bone defects are automatically identified, the position of tooth root tips is automatically located, and the local bone absorption rate is calculated.

Systems and methods for automatically segmenting and detecting a menstrual cycle phase in ultrasound images of anatomical structures that change over a patient menstrual cycle are provided. The method includes acquiring, by an ultrasound probe of an ultrasound system, an ultrasound image of a region of interest having an anatomical structure that changes over a patient menstrual cycle. The method includes automatically segmenting, by at least one processor of the ultrasound system, an anatomical structure depicted in the ultrasound image. The method includes automatically predicting, by the at least one processor, a menstrual cycle phase based on the segmentation of the anatomical structure. The method includes causing, by the at least one processor, a display system to present at least one rendering of the segmented anatomical structure and the predicted menstrual cycle phase.